A shield segment assembly error quantification and position prediction method, system and device

Abstract

The application relates to a shield tunnel segment assembling error quantification and segment position prediction method, and relates to the technical field of shield subway tunnel construction. The position deviation of a point B is calculated by combining the horizontal and vertical coordinate positions of the position deviation of the point B with the coordinate data of a center position A of a segment ring at a position of a laser target and the distance of the center position B of the outermost segment ring and the actually measured position of the laser target. Then, the measured data and the actual position data of the segment are compared, and the center position data of a segment assembling ring in different tunneling strata and error rules are calculated. The error is corrected by comparing the data of a trial tunneling section, a segment position prediction model is established by using an artificial neural network algorithm, and the relative position of a next segment ring is determined. The application can judge the relative position relationship between the segment position and the tunnel line type during tunneling, and can predict the relative position of the next segment ring. The application can significantly reduce the error in the segment assembling process, and can guarantee the assembling quality and forming quality of the shield tunnel segment.

Description

Technical Field

[0001] This invention relates to the field of shield tunnel construction technology, specifically to a method, system, and device for quantifying and predicting the location of shield tunnel segment assembly errors. Background Technology

[0002] The core of tunnel construction quality lies in the precise assembly of tunnel segments, which directly determines the integrity and durability of the tunnel structure. In traditional shield tunnel construction, segment assembly usually relies on propulsion cylinders, tunneling parameters, and the positioning system of the tunnel boring machine. These methods are often passive and lack real-time capability.

[0003] The tunnel boring machine (TBM) has a complex overall structure, with a large distribution of propulsion cylinders and complex connections between the cylinders and their corresponding sensors, making it easy for segment assembly errors to occur. Furthermore, because it is manually operated, there is a degree of subjectivity involved, and the segment positions cannot be determined before assembly. Misalignments caused during tunneling and assembly can easily lead to segment alignment quality problems, such as misalignment, leakage, damage, and cracking. These problems not only increase the difficulty and cost of later repairs but also pose a serious threat to construction safety and project quality.

[0004] Furthermore, the significant differences in the physical and mechanical properties of different geological strata have a substantial impact on segment assembly errors. However, current technologies lack a mechanism to effectively combine geological parameters with error calculations, resulting in limited accuracy in error calculations. Additionally, there is no robust early warning and automatic adjustment mechanism when errors occur, making it difficult to respond quickly and address error issues.

[0005] The above analysis reveals numerous challenges in using existing technologies for shield tunnel construction: low construction efficiency, high costs, and difficulty in guaranteeing quality. Especially in complex geological formations, accurately controlling the assembly position of tunnel segments, reducing errors, and improving construction precision have become pressing technical problems that need to be solved. Summary of the Invention

[0006] The technical problem to be solved by the present invention is that the existing technology has low construction accuracy and inaccurate grasp of the assembly position of tunnel segments. The purpose is to provide a method, system and device for quantifying the assembly error and predicting the position of tunnel segments. By adopting the corresponding technical solution, the assembly of tunnel segments can be guided more accurately, and the beneficial effects of accurate prediction, high construction efficiency and high construction quality are achieved.

[0007] This invention is achieved through the following technical solution:

[0008] The first invention provides a method for quantifying and predicting the location of shield tunnel segment assembly errors, comprising,

[0009] S1. The shield machine's attitude is monitored in real time by a laser target in the guidance system during the tunneling process;

[0010] S2. Obtain the coordinate data (xA, yA) of the center position A of the tube ring where the laser target is located.

[0011] S3. Obtain the distance L between the laser target position and the center point B of the outermost tube ring;

[0012] S4. Calculate the positional deviation (xB, yB) of point B by combining the horizontal trend (kx) and vertical trend (ky) measured by the tunnel boring machine. Compare the calculated data with the positional data (x1, y1) of the segment remeasured by the measurement group to obtain the error data (Δx, Δy); where: xB=xA+L×kx, yB=yA+L×ky; Δx=|x1-xB|; Δy=|y1-yB|;

[0013] S5. Summarize the data based on multiple (Δx, Δy) data to obtain the center position data and error variation law of the segment assembly ring under different tunneling strata; S6. Establish a segment position prediction model through artificial neural network algorithm, and continuously track the attitude data of the segments throughout the tunneling process to optimize the prediction model and determine the relative position of the next ring segment.

[0014] Furthermore, in this invention, in step S4, a dynamic correction coefficient (λ) for formation parameters is introduced. This dynamic correction coefficient (λ) is determined based on the hardness, water content, and porosity of the formation and is incorporated into the calculation of the position deviation at point B. The corrected formula for calculating the coordinates of point B is: xB = xA + L × kx × λ, yB = yA + L × ky × λ. In step S5, an error warning mechanism is established. When the error data (Δx, Δy) exceeds a preset threshold, an early warning is automatically triggered and an error adjustment scheme is generated. The preset threshold is dynamically set according to the error variation law of different formations.

[0015] Furthermore, in this invention, the hardness, water content, and porosity data of different formations are obtained through indoor tests. The weight of each parameter is determined by the analytic hierarchy process (AHP), and the dynamic correction coefficient (λ) of the formation parameters is obtained by weighted calculation. The specific calculation formula is λ=α×H+β×W+γ×P, where H is the formation hardness parameter, W is the formation water content parameter, P is the formation porosity parameter, α, β, and γ are the weights of each parameter, and α+β+γ=1.

[0016] Furthermore, in this invention, the dynamic setting method of the preset threshold in the above-mentioned error warning mechanism is as follows: based on the error change law under different strata, the initial threshold corresponding to each stratum is determined by statistical analysis method. During the tunneling process, the threshold is dynamically adjusted by exponential smoothing method according to the degree of deviation between the real-time error data and the initial threshold.

[0017] Furthermore, in this invention, in step S1 above, the data from the laser target is used to monitor the shield's attitude in real time, including deviations in the horizontal and vertical directions, and in step S6, the artificial neural network algorithm is neuron Ai in the artificial neural network.

[0018] Furthermore, in this invention, the aforementioned artificial neural network algorithm includes an input layer, a hidden layer, and an output layer;

[0019] The input layer vector is represented as X = (x1, x2, ..., xk, ... xn) T, the hidden layer output vector is represented as Y = (y1, y2, ..., yk, ... yn) T, and the output layer output vector is represented as O = (o1, o2, ..., ok, ... on) T. The weight matrices between the input layer and the hidden layer are represented as V = (v1, v2, ..., vk, ... vn) and W = (ω1, ω2, ..., ωk, ... ωn), respectively, where vk and ωk represent the weight values ​​of the k-th neuron.

[0020] The mathematical formula for the hidden layer is: ,

[0021] Where vij is the connection weight between the i-th neuron in the input layer and the j-th neuron in the hidden layer, xi is the input value of the i-th neuron in the input layer, yj is the output value of the j-th neuron in the hidden layer, and j is the number of the hidden layer neuron.

[0022] The mathematical formula for the output layer is: ,

[0023] Where ωjk is the connection weight between the j-th neuron in the hidden layer and the k-th neuron in the output layer, yj is the output value of the j-th neuron in the hidden layer, ok is the output value of the k-th neuron in the output layer, and k is the number of the neuron in the output layer; where f1(x) and f2(x) represent the transfer functions of the output layer and the hidden layer, respectively.

[0024] Furthermore, in this invention, the aforementioned input layer receives input information (x0…xn) from the upper-level neurons and processes the input information in the following manner before outputting Oi:

[0025] ,

[0026] Where Ti represents the threshold of artificial neuron Ai, ωki represents the connection weight between neurons Ak and Ai, and f(x) is the neuron transfer function.

[0027] Secondly, the present invention also provides a shield tunnel segment assembly error quantification and location prediction system, which employs the aforementioned shield tunnel segment assembly error quantification and location prediction method, including:

[0028] (1) Attitude monitoring module, used to monitor the attitude of the shield in real time through the laser target of the guide system during the tunneling process;

[0029] (2) Coordinate calculation module, used to calculate the coordinate data of the center position A of the tube ring where the laser target is located;

[0030] (3) Distance measurement module, used to obtain the distance between the laser target position and the center point B of the outermost tube ring;

[0031] (4) Error calculation module, used to calculate the position deviation of point B by combining the horizontal and vertical trends measured by the tunnel boring machine, and compare it with the position data of the re-measured segments to obtain error data;

[0032] (5) Pattern summarization module, used to summarize multiple error data to obtain the center position data and error variation pattern of the segment assembly ring under different tunneling strata;

[0033] (6) Prediction model module, used to establish a segment position prediction model through artificial neural network algorithm, and continuously track the segment attitude data to optimize the prediction model;

[0034] (7) Formation parameter correction module, used to obtain formation hardness, water content and porosity data, calculate formation parameter dynamic correction coefficient (λ), and apply it to the correction calculation of the position deviation of point B;

[0035] (8) Early warning and adjustment module, used to set the preset error threshold, issue an early warning when the error data exceeds the threshold, and generate an error adjustment plan.

[0036] Furthermore, in this invention, the above also includes a database for storing the shield machine's attitude monitoring data, coordinate calculation data, distance measurement data, error calculation data, pattern summary data, prediction model data, stratum parameter data, and early warning adjustment data.

[0037] Thirdly, the present invention also provides a device for quantifying and predicting the assembly error of tunnel segments, which includes a system for quantifying and predicting the assembly error of tunnel segments.

[0038] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0039] This invention introduces a dynamic correction coefficient (λ) for formation parameters, which fully considers the influence of different formations on segment assembly errors. By incorporating key parameters such as formation hardness, water content, and porosity into the calculation of the position deviation at point B, the error calculation becomes more accurate, overcoming the problem of inaccurate error calculation caused by neglecting formation differences in traditional methods.

[0040] This invention establishes an error early warning mechanism that can issue timely warnings when errors exceed a preset threshold and automatically generate adjustment plans. This facilitates rapid corrective measures by construction personnel, preventing the accumulation of errors from severely impacting tunnel construction quality. Furthermore, the preset threshold is dynamically set based on the error variation patterns of different geological strata, improving the targeting and accuracy of the early warnings.

[0041] This invention utilizes a laser-guided system to monitor the shield tunnel's attitude in real time. During shield tunnel excavation, it can effectively calculate and analyze segment assembly errors and their variation patterns. Furthermore, it can determine the assembly position of the next segment using formulas and real-time monitoring data. By reducing segment assembly errors and optimizing segment selection, construction efficiency can be significantly improved, construction costs reduced, and the number of personnel required decreased. Precise segment assembly position prediction and error control effectively ensure the overall tunnel's forming quality, reducing the need for subsequent repairs and maintenance.

[0042] Compared to traditional prediction methods, this invention applies an artificial neural network algorithm to establish a segment location prediction model, which can be continuously optimized based on real-time monitoring data to improve prediction accuracy. As the tunnel boring machine progresses, the model can continuously absorb new error data, learn and adapt, and correct and optimize subsequent prediction data. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0044] Figure 1 This is a flowchart of the method in Embodiment 1 of the present invention;

[0045] Figure 2 This is a schematic diagram of the ideal structure in which the center point B of the outermost segment ring of the present invention coincides with the center point A of the segment ring where the laser target is located.

[0046] Figure 3 This is a schematic diagram of an ideal structure in which the center point B of the outermost segment ring of the present invention does not coincide with the center point A of the segment ring where the laser target is located;

[0047] Figure 4 This is a schematic diagram of the geometric relationship of the positional deviation of point B in the X direction of the present invention;

[0048] Figure 5 This is a schematic diagram of the geometric relationship of the positional deviation of point B in the Y direction of the present invention;

[0049] Figure 6 This is a diagram showing the horizontal and vertical deviations of the tunnel lining segments of the present invention as the mileage increases;

[0050] Figure 7 This is a schematic diagram illustrating the artificial neuron's reception and processing of information according to the present invention.

[0051] Figure 8 This is a comparison diagram of the training and prediction structures of this invention;

[0052] Figure 9 This is a flowchart illustrating the calculation of dynamic correction coefficients for formation parameters in Embodiment 2 of the present invention.

[0053] Figure 10 This is a flowchart of the error warning mechanism in Embodiment 2 of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments and accompanying drawings. The illustrative embodiments and descriptions of this invention are for explanation only and are not intended to limit the invention. The following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0055] Example 1: The overall process of the present invention is as follows Figure 1 As shown, in the early stage of the tunnel boring machine (TBM) excavation, the attitude data of the center position A of the segment ring at the location of the laser target in the TBM's guidance system is monitored in real time. Then, based on the distance L between the laser target position and the center point B of the outermost segment ring measured on-site, and combined with the horizontal and vertical coordinates of the TBM itself, the position deviation of point B is calculated. The measured data is compared with the actual position data of the segment rings, and the center position data and error patterns of the segment assembly rings under different excavation strata are summarized. The error is corrected by comparing the error of the test excavation section. The error data and variation patterns of segment assembly under different strata can be obtained, and the relative position of the next ring segment can be determined, providing a basis for segment selection.

[0056] like Figure 2As shown, the left side represents the excavation section of the tunnel boring machine (TBM). During TBM construction, the ideal positional relationship between the outermost segment ring center point B and the segment ring center point A (located at the laser target) is when they coincide. However, in actual construction, the TBM needs to fit the design axis during excavation. The shield tail and the front shield often form a "V" shape, meaning the shield tail centerline does not coincide with the segment ring center point A (located at the laser target). Furthermore, during excavation, the TBM's correction work to fit the design axis results in random center positions for each ring. Therefore, in actual construction, it is necessary to calculate the positional deviation of point B by combining the TBM's own horizontal and vertical coordinates to determine the horizontal or vertical positional relationship between the segment and the shield tail.

[0057] like Figure 3 As shown, the center point B of the outermost segment ring does not coincide with the center point A of the segment ring where the laser target is located. At this time, it is necessary to analyze the attitude data of the center point A of the segment ring where the laser target is located, and measure the distance L between the laser target position and the center point B of the outermost segment ring. Then, the position deviation of the center point B of the outermost segment ring can be obtained by the trend of the horizontal and vertical deviations.

[0058] For example: After a certain ring is advanced, according to laser measurement data, the deviation of the front shield in the horizontal position is 0mm, the deviation of the middle shield is 8mm, the deviation of the rear shield is 14mm, and the trend is -2.

[0059] Then, the distance L between the laser target position and the center point B of the outermost tube ring, as measured on-site, is 10mm.

[0060] Depend on Figure 4 From the given information, we can see that xB = xA + L × kx = 14 + 10 × (-2) = -6mm, which means that the positional deviation of the outermost tube ring center point B relative to the tube ring center point A where the laser target is located is 20mm.

[0061] After a certain ring was advanced, according to laser measurement data, the deviation of the front shield in the vertical position was -6mm, the deviation of the middle shield was -24mm, the deviation of the rear shield was -30mm, and the trend was 4.

[0062] Similarly, from Figure 5 It can be seen that yB=yA+L×ky=-30+10×(4)=10mm, which means that the position deviation of the outermost tube ring center point B relative to the tube ring center point A where the laser target is located is 40mm.

[0063] Based on the above calculation process, the positional deviation of point B is (-6, 10).

[0064] The measurement group re-measured the position data of the segment (-10, 20).

[0065] The calculated data is compared with the position data (x1, y1) of the re-measured pipe segment by the measurement group to obtain the error data (Δx, Δy).

[0066] The formulas for quantifying error data are Δx = |x1 - xB|; Δy = |y1 - yB|.

[0067] Δx = |x1 - xB| = |-10 + 6| = 4 mm.

[0068] Δy=|y1-yB|=|10-20|=10mm.

[0069] When this invention was applied to a certain bidding section, the collected data were as follows: Figure 6 As shown, the horizontal and vertical deviations (Δx, Δy) of the shield tunneling attitude data as the mileage increases are calculated according to the above method, and the data are statistically analyzed using an EXCEL spreadsheet.

[0070] like Figure 7 As shown, in an artificial neural network, neuron Ai receives input information (x0…xn) from its parent neurons, processes it, and then outputs information. ,in: .

[0071] Where Ti and Tj represent the thresholds of artificial neurons Ai and Aj, ωki and ωij represent the connection relationships between neurons Ak and Ai, and between neurons Ai and Aj (here referring to the relative weights of the structural defect index), and f(x) is the neuron transfer function.

[0072] Through special connections between artificial neurons, artificial neural networks composed of multi-layered structures can be constructed. These networks, through input, hidden, and output layers, can solve complex real-world problems influenced by multiple factors. The input layer vector can be represented as X = (x1, x2, ..., xk, ... xn)T, the hidden layer output vector as Y = (y1, y2, ..., yk, ... yn)T, the output layer output vector as O = (o1, o2, ..., ok, ... on)T, and the desired output vector as P = (p1, p2, ..., pk, ... pn)T. The weight matrices between the input and hidden layers are represented by V = (v1, v2, ..., vk, ... vn) and W = (ω1, ω2, ..., ωk, ... ωn), where vk and ωk represent the weight values ​​of the k-th neuron.

[0073] The mathematical formula for the output layer is: .

[0074] The mathematical formula for the hidden layer is: .

[0075] Where f1(x) and f2(x) represent the output layer and hidden layer transfer functions, also known as activation functions, respectively.

[0076] A formula for calculating the position of tunnel segments is established using a programming language, and the relative position of the next ring of tunnel segments can be determined by real-time monitoring data from the guidance system.

[0077] The top 60 sets of data on the quantization error of the above-mentioned segments were selected and substituted into the software Matlab for calculation.

[0078] like Figure 8 As shown, the comparison of the prediction results of the training set reflects that the difference between the true value and the predicted value is not significant, proving that the prediction model is suitable for segment location prediction.

[0079] Meanwhile, as the tunnel boring machine advances, new error data will be generated. After being added to the database, the original model will be continuously optimized in conjunction with the algorithm to correct subsequent prediction data.

[0080] After implementing the above calculation method, the positional relationship and error between the shield tunnel segment alignment points and the segments in the vertical and horizontal directions can be calculated. Combining the segment position deviation patterns, the segment position prediction model, and subsequent correction data, the assembly position of the next ring of segments can be determined, and the segment selection can be optimized.

[0081] Example 2: The method for quantifying and predicting the location of shield tunnel segment assembly errors in Example 2 is a further optimization based on Example 1, which introduces dynamic correction coefficients for geological parameters and establishes an error early warning mechanism.

[0082] The calculation process for the dynamic correction coefficient (λ) of formation parameters is as follows: Figure 9 As shown, firstly, the hardness, water content and porosity data of the formation are obtained through indoor tests. Then, the weight of each parameter is determined by the analytic hierarchy process. Finally, the dynamic correction coefficient (λ) of the formation parameters is calculated according to the formula λ=α×H+β×W+γ×P.

[0083] For example, for a certain sandy soil layer, H=0.6, W=0.3, P=0.2 are measured through experiments. Using the analytic hierarchy process (AHP), α=0.5, β=0.3, γ=0.2 are determined. Then λ=0.5×0.6+0.3×0.3+0.2×0.2=0.3+0.09+0.04=0.43.

[0084] The dynamic correction factor (λ) for formation parameters is applied to the calculation of the position deviation at point B.

[0085] For example: After a certain ring is advanced, according to laser measurement data, the horizontal deviation of the front shield is 0mm, the middle shield deviation is 8mm, and the rear shield deviation is 14mm, with a trend of -2. The distance L between the laser target position and the center point B of the outermost segment ring, as measured on-site, is 10mm, and the dynamic correction coefficient for formation parameters is λ=0.43.

[0086] Therefore, the corrected xB = xA + L × kx × λ = 14 + 10 × (-2) × 0.43 = 14 - 8.6 = 5.4 mm.

[0087] After a certain ring was advanced, according to laser measurement data, the vertical deviation of the front shield was -6mm, the middle shield deviation was -24mm, and the rear shield deviation was -30mm, with a trend of 4, L=10mm, and λ=0.43.

[0088] The corrected yB = yA + L × ky × λ = -30 + 10 × 4 × 0.43 = -30 + 17.2 = -12.8 mm.

[0089] Based on the above calculation process, the corrected positional deviation of point B is (5.4, -12.8).

[0090] The measurement group remeasured the position data of the pipe segment as (5.0, -13.0).

[0091] The calculated data is compared with the position data (x1, y1) of the re-measured segments by the measurement group to obtain the error data (Δx, Δy).

[0092] The formulas for quantizing error data are Δx = |x1 - xB|; Δy = |y1 - yB|.

[0093] Δx = |5.0 - 5.4| = 0.4 mm

[0094] Δy = |-13.0 - (-12.8)| = 0.2 mm

[0095] It is evident that by introducing dynamic correction coefficients for formation parameters, the error data is smaller and the calculation results are more accurate.

[0096] The workflow of the error warning mechanism is as follows: Figure 10 As shown, firstly, based on the error variation patterns under different strata, statistical analysis methods are used to determine the initial thresholds corresponding to each stratum. For example, for clay layers, the initial threshold in the horizontal direction is set to 5 mm, and the initial threshold in the vertical direction is set to 4 mm. During tunneling, error data is monitored in real time. When the horizontal error Δx = 6 mm, exceeding the initial threshold of 5 mm, the threshold is dynamically adjusted using the exponential smoothing method. The adjusted horizontal threshold may be 5.5 mm, and an early warning is issued.

[0097] Example 3: This invention provides a shield tunnel segment assembly error quantification and location prediction system, which adopts the shield tunnel segment assembly error quantification and location prediction method of Example 2, including:

[0098] (1) Attitude monitoring module, used to monitor the attitude of the shield in real time through the laser target of the guide system during the tunneling process;

[0099] (2) Coordinate calculation module, used to calculate the coordinate data of the center position A of the tube ring where the laser target is located;

[0100] (3) Distance measurement module, used to obtain the distance between the laser target position and the center point B of the outermost tube ring;

[0101] (4) Error calculation module, used to calculate the position deviation of point B by combining the horizontal and vertical trends measured by the tunnel boring machine, and compare it with the position data of the re-measured segments to obtain error data;

[0102] (5) Pattern summarization module, used to summarize multiple error data to obtain the center position data and error variation pattern of the segment assembly ring under different tunneling strata;

[0103] (6) Prediction model module, used to establish a segment position prediction model through artificial neural network algorithm, and continuously track the segment attitude data to optimize the prediction model;

[0104] (7) Formation parameter correction module, used to obtain formation hardness, water content and porosity data, calculate formation parameter dynamic correction coefficient (λ), and apply it to the correction calculation of the position deviation of point B;

[0105] (8) Early warning and adjustment module, used to set the preset error threshold, issue an early warning when the error data exceeds the threshold, and generate an error adjustment plan.

[0106] The shield tunnel segment assembly error quantification and location prediction system also includes a database for storing shield machine attitude monitoring data, coordinate calculation data, distance measurement data, error calculation data, pattern summary data, prediction model data, stratum parameter data, and early warning adjustment data.

[0107] Example 4: The present invention provides a shield tunnel segment assembly error quantification and location prediction device, which adopts the shield tunnel segment assembly error quantification and location prediction system of Example 2.

[0108] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for quantifying and predicting the location of shield tunnel segment assembly errors, characterized in that, include, S1. The shield machine's attitude is monitored in real time by a laser target in the guidance system during the tunneling process; S2. Obtain the coordinate data (xA, yA) of the center position A of the tube ring where the laser target is located; S3. Obtain the distance L between the laser target position and the center point B of the outermost tube ring; S4. Calculate the position deviation (xB, yB) of point B by combining the horizontal trend kx and vertical trend ky measured by the tunnel boring machine. Compare the calculated data with the position data (x1, y1) of the segment remeasured by the measurement group to obtain the error data (Δx, Δy); where: xB=xA+L×kx, yB=yA+L×ky; Δx=|x1-xB|; Δy=|y1-yB|; S5. Summarize multiple (Δx, Δy) data to obtain the center position data and error variation law of the segment assembly ring under different tunneling strata; S6. Establish a segment position prediction model through artificial neural network algorithm, and continuously track the attitude data of the segments throughout the tunneling process to optimize the prediction model and determine the relative position of the next ring of segments. In step S4, a dynamic correction coefficient λ for formation parameters is introduced. The dynamic correction coefficient λ for formation parameters is determined based on the hardness, water content and porosity of the formation. It is incorporated into the calculation of the position deviation of point B. The corrected formula for calculating the coordinates of point B is: xB=xA+L×kx×λ,yB=yA+L×ky×λ。 Furthermore, in step S5, an error early warning mechanism is established. When the error data (Δx, Δy) exceeds the preset threshold, an early warning is automatically triggered and an error adjustment scheme is generated. The preset threshold is dynamically set according to the error change pattern of different strata.

2. The method for quantifying and predicting the location of shield tunnel segment assembly errors according to claim 1, characterized in that, Data on hardness, water content, and porosity of different formations were obtained through indoor experiments. The weight of each parameter was determined by the analytic hierarchy process (AHP), and the dynamic correction coefficient λ of the formation parameters was obtained by weighted calculation. The specific calculation formula is λ=α×H+β×W+γ×P, where H is the formation hardness parameter, W is the formation water content parameter, P is the formation porosity parameter, α, β, and γ are the weights of each parameter, and α+β+γ=1.

3. The method for quantifying and predicting the location of shield tunnel segment assembly errors according to claim 1, characterized in that, In the aforementioned error warning mechanism, the dynamic setting method for the preset threshold is as follows: based on the error variation patterns under different strata, the initial threshold corresponding to each stratum is determined by statistical analysis. During the tunneling process, the threshold is dynamically adjusted by exponential smoothing based on the degree of deviation between the real-time error data and the initial threshold.

4. The method for quantifying and predicting the location of shield tunnel segment assembly errors according to claim 1, characterized in that, In step S1, the data from the laser target is used to monitor the shield's attitude in real time, including deviations in the horizontal and vertical directions. In step S6, the artificial neural network algorithm is the neuron Ai in the artificial neural network.

5. The method for quantifying and predicting the location of shield tunnel segment assembly errors according to claim 4, characterized in that, Artificial neural network algorithms consist of an input layer, hidden layers, and an output layer; The input layer vector is represented as X = (x1, x2, ..., xk, ... xn)T, the hidden layer output vector is represented as Y = (y1, y2, ..., yk, ... yn)T, and the output layer output vector is represented as O = (o1, o2, ..., ok, ... on)T. The weight matrices between the input layer and the hidden layer are represented by V = (v1, v2, ..., vk, ... vn) and W = (ω1, ω2, ..., ωk, ... ωn), respectively, where vk and ωk represent the weight values ​​of the k-th neuron. The mathematical formula for the hidden layer is: Where vij is the connection weight between the i-th neuron in the input layer and the j-th neuron in the hidden layer, xi is the input value of the i-th neuron in the input layer, yj is the output value of the j-th neuron in the hidden layer, and j is the number of the hidden layer neuron. The mathematical formula for the output layer is: Where ωjk is the connection weight between the j-th neuron in the hidden layer and the k-th neuron in the output layer, yj is the output value of the j-th neuron in the hidden layer, ok is the output value of the k-th neuron in the output layer, and k is the number of the neuron in the output layer. Where f1(x) and f2(x) represent the output layer and hidden layer transfer functions, respectively.

6. The method for quantifying and predicting the location of shield tunnel segment assembly errors according to claim 5, characterized in that, The input layer receives input information (x0…xn) from the upper-level neurons and processes the input information in the following way before outputting Oi: Where Ti represents the threshold of artificial neuron Ai, ωki represents the connection weight between neurons Ak and Ai, and f(x) is the neuron transfer function.

7. A system for quantifying and predicting the location of shield tunnel segment assembly errors, characterized in that: The method for quantifying and predicting the location of shield tunnel segment assembly errors as described in claim 1 includes: (1) Attitude monitoring module, used to monitor the attitude of the shield in real time through the laser target of the guide system during the tunneling process; (2) Coordinate calculation module, used to calculate the coordinate data of the center position A of the tube ring where the laser target is located; (3) Distance measurement module, used to obtain the distance between the laser target position and the center point B of the outermost tube ring; (4) Error calculation module, used to calculate the position deviation of point B by combining the horizontal and vertical trends measured by the tunnel boring machine, and compare it with the position data of the re-measured segments to obtain error data; (5) Pattern summarization module, which is used to summarize multiple error data to obtain the center position data and error variation law of the segment assembly ring under different tunneling strata; (6) Prediction model module, which is used to establish a segment position prediction model through artificial neural network algorithm and continuously track the segment attitude data to optimize the prediction model; (7) Formation parameter correction module, used to obtain formation hardness, water content and porosity data, calculate formation parameter dynamic correction coefficient λ, and apply it to the correction calculation of the position deviation of point B; (8) Early warning and adjustment module, used to set the preset error threshold, issue an early warning when the error data exceeds the threshold, and generate an error adjustment plan.

8. The shield tunnel segment assembly error quantification and location prediction system according to claim 7, characterized in that: It also includes a database for storing attitude monitoring data, coordinate calculation data, distance measurement data, error calculation data, pattern summary data, prediction model data, geological parameter data, and early warning adjustment data of the tunnel boring machine.

9. A device for quantifying and predicting the location of shield tunnel segment assembly errors, characterized in that, It includes the shield tunnel segment assembly error quantification and location prediction system as described in claim 8.

Images